Low evidence for ringdown overtone in GW150914 when marginalizing over time and sky location uncertainty

Tests of the no-hair theorem using astrophysical black holes involve the detection of at least two quasinormal modes (QNMs) in the gravitational wave emitted by a perturbed black hole. A detection of two modes—the dominant, $(\ell ,m,n)=(2,2,0)$, mode and its first overtone, the (2, 2, 1) mode—in the postmerger signal of the binary black hole merger GW150914 was claimed by Isi et al. [Testing the no-hair theorem with GW150914, Phys. Rev. Lett. 123, 111102 (2019)], with further evidence provided by Isi and Farr [Revisiting the ringdown of GW150914 (2022)]. However, Cotesta et al. [Analysis of ringdown overtones in GW150914, Phys. Rev. Lett. 129, 111102 (2022)] disputed this claim, finding that evidence for the overtone only appeared if the signal was analyzed before merger, when a QNM description of the signal is not valid. Because of technical challenges, both of these analyses fixed the merger time and sky location of GW150914 when estimating the evidence for the overtone. At least some of the contention can be attributed to fixing these parameters. Here, we surmount this difficulty and fully marginalize over merger time and sky-location uncertainty while doing a postmerger QNM analysis of GW150914. We find that marginalizing over all parameters yields low evidence for the presence of the overtone, with a Bayes factor of $1.10\pm 0.03$ in favor of a QNM model with the overtone versus one without. The arrival time uncertainty of GW150914 is too large to definitively claim detection of the (2, 2, 1) mode.